GB2313224A

GB2313224A - Ferroelectric liquid crystal device

Info

Publication number: GB2313224A
Application number: GB9610313A
Authority: GB
Inventors: Mitsuhiro Shigeta; Jonathan Rennie Hughes; Paul Bonnett; Edward Peter Raynes; Akira Tagawa; Michael John Towler; Harry Garth Walton; David Charles Scattergood
Original assignee: UK Secretary of State for Defence; Sharp Corp
Current assignee: UK Secretary of State for Defence; Sharp Corp
Priority date: 1996-05-17
Filing date: 1996-05-17
Publication date: 1997-11-19
Also published as: EP0809231A3; US6215533B1; JPH1054974A; EP0809231A2; GB9610313D0

Description

M&C Folio: 230P74506 2313224 Ferroelectric Liquid Qystal Device The

present invention relates to a ferroelectric liquid crystal device such as a large area flat panel display comprising a driving arrangement for reducing adverse effects caused by non-uniform heating of the device. The invention further relates to a driving arrangement for a ferroelectric liquid crystal array device and to a method of driving a ferroelectric liquid crystal array device.

Ferroelectric liquid crystal materials are of important application to flat panel liquid crystal array devices because of their high switching speed and bistability. Unlike supertwist nematic liquid crystal devices, for example, the pixels of such a device will remain in a particular state without continued application of a particular drive voltage. In a large area panel display device which has to be addressed by multiplexing this is a significant advantage. Ferroelectric liquid crystal arrays and a driving scheme therefor are described in 'The JOERS/Alvey Ferroelectric Multiplexing Scheme' published in Ferroelectrics, 1991, Vol. 122 pages 63 to 79. In such driving schemes a liquid crystal array has a first and second set of driving electrodes arranged at right angles to each other defining a matrix. A plurality of pixels are defined at the intersection of an electrode from the first plurality and an electrode from the second plurality. However, by the very nature of this layout, it is not possible to address each pixel individually. The type of addressing scheme used most commonly applies a strobe signal in sequence to one of the sets of electrodes (referred to hereafter as the row electrodes) while applying the relevant data signals for the currently-strobed row to the second set of electrodes (hereafter referred to as the column electrodes).

One consequence of such a scheme is that the signals applied to the column electrodes are applied across every pixel in the respective column, even though only one pixel is actually being addressed at any one time. In a ferroelectric display it is not feasible to 2 remove such signals (for example by open-circuiting the non-strobed row column electrodes) because they are required to apply an AC stabilisation signal to the cells of the array. Such a signal prevents the liquid crystal molecules in the cells relaxing to a position which has an unfavourable optical performance. These signals, however, are continually applied at a high frequency to every column electrode to drive a capacitive load comprising the cells of the device. The column electrodes generally comprise transparent indium tin oxide (ITO) tracks which have a certain resistance so the charging and discharging of the cells dissipates power in these tracks which heats the device.

The temperature of the device is particularly critical in a ferroelectric liquid crystal array device because of the large temperature sensitivity of ferroelectric materials themselves. To some extent effects of global temperature changes to the device can be compensated for in the addressing waveforms. For example changes in the switching speed (operating window) can be compensated for by changing the shape or amplitude of the strobe voltage, whilst changes in memory angle can be compensated for by changing the amplitude of the column (data) waveforms. However, the prior art drive schemes such as the one described in the reference above, apply rectangular waves to the column electrodes to drive the device and these waveforms have a rich harmonic content including substantial frequency components at high multiples of the fundamental frequency. Since each column of the array appears as a distributed RC ladder to the driving circuitry, these higher harmonics of the driving waveform are attenuated heavily by the device and the highest attenuation occurs at the driven end of the column electrodes, in other words at the edge of the device. This causes nonuniform heating of the panel that cannot be compensated by adjusting the row or column signals (since they clearly apply to all of the cells in a column). The consequence of this is variations in contrast or colour over the array display device (or, in extreme cases failure to switch when addressed) which is unnacceptable. Liquid crystal devices based on nematic liquid crystal phases do not suffer from these problems because of their higher tolerance of temperature variations.

3 It is an object of the present invention to ameliorate the above problem in ferroelectric liquid crystal devices.

It is a further object of the invention to provide a novel driving arrangement for a ferroelectric liquid crystal array device and to provide a novel method of driving such a device.

According to a first aspect of the present invention there is provided a ferroelectric liquid crystal device as set out in appended claim 1.

According to a second aspect of the present invention there is provided a driving arrangement for a ferroelectric liquid crystal array device as set out in appended claim 14.

According to a third aspect of the present invention there is provided a method of driving a ferroelectric liquid crystal array device as set out in appended claim 15.

The present invention is based upon the realisation that the non-uniform heating of a ferroelectric liquid crystal device as described above can be reduced considerably by driving the column electrodes with a signal that is substantially lower in harmonic content than the square wave style driving waveforms of the prior art. Particular nonrectangular waveforms are of interest such as sinusiodal waveforms, triangular waveforms and trapezoidal waveforms. The sinusoidal waveform clearly has the lowest harmonic content of the three: ideally being zero above the fundamental frequency. However, the higher harmonic content of the other two waveforms is low and these waveforms have the advantage that they can generally be provided with simpler circuit arrangements than can a suitable sinusiodal waveform.

The present invention also has these further advantages: (i) the total power dissipation in the panel is reduced, making it less susceptible to overall temperature-related effects, 4 (ii) the attenuation of the column waveforms along the ITO tracks is reduced which means that the waveform applied to the cells most distant from the drive circuitry is less distorted and so the switching of the pixels is more reliable, and (iii) the power dissipated is less dependent upon the pixel pattern (in other words the image displayed) whichTurther reduces undesirable temperature-related effects.

The strobe waveform applied to the row electrodes may also be provided to be a reduced-harmonic waveform but this is not as desirable as providing reduced-harmonc column waveforms. The strobe waveform is applied to each row only for the time that the row is being addressed which in a large panel array is a very short period of time. Thus the heating effects of this waveform are not nearly as significant as those of the column waveforms. It is not generally worthwhile to provide the more sophisticated drive crcuitry for the row driving arrangement. In addition, if the strobe waveform is of a sinusoidal shape then the operating window may be shifted to higher peak voltages than for that of a rectangular strobe waveform.

As mentioned above, however, it is not necessary to apply a pure sinusoid to the column waveforms to obtain the advantages of the present invention. A range of waveforms may be applied that have a reduced higher-hannonic content compared with a square wave. Waveforms that provide the advantages of the invention can be defined in terms of their harmonic content and their interaction with the other parameters of the liquid crystal device. For example, the waveform may be defined in terms of its power dissipation when applied to the panel or the distortion of the waveform along the length of the electrodes.

The present invention will now be described, by way of example, with reference to the accompanying drawings, in which:

Figure 1 shows a block schematic diagram of a ferroelectric liquid crystal flat panel display and driving circuitry, Figure 2 shows a lumped equivalent circuit of a ferroelectric liquid crystal device, Figure 3 shows a distributed equivalent circuit of a ferroelectric liquid crystal device, Figure 4 shows three possible waveforms which may be applied to the column electrodes of a liquid crystal array in accordance with the invention, Figure 5 shows a block schematic diagram of a driving arrangement for applying signals to the column electrodes in accordance with the present invention, Figure 6 shows strobe and column (data) waveforms for application to a panel in accordance with the present invention, Figure 7 shows a graph of switching time against applied voltage (a so- called -CV graph) for a ferroelectric liquid crystal display in accordance with the invention, Figure 8 shows a pair of prior art data waveforms for illustrating a further advantage of the invention,

Figure 9 shows the equivalent waveforms to those of figure 8 for the present invention, Figure 10 is a block schematic diagram of a rig for testing the present invention, and Figure 11 shows a graph of temperature variation against distance from the edge of a ferroelectric liquid crystal array panel for a prior art drive scheme and a drive scheme in accordance with the present invention.

Figure 1 shows a ferroelectric liquid crystal array device 10 comprising a first transparent wall 12 and a second transparent wall 20 spaced apart from the first wall by known means such as spacer beads (not shown). The wall 12 carries a plurality of electrodes 16 of transparent tin oxide on that surface of the wall that faces the second wall 20. The electrodes 16 are arranged parallel to one another and each extend between a first edge of the wall 12 and a second edge at which an electrical connector 14 is arranged to connect each electrode to a column driver 18. The wall 20 carries a plurality of transparent electrodes 22 also arranged in parallel with one another but at right angles to the electrodes 16 on the first wall. The electrodes 22 extend from a first edge of the wall 20 to a second edge at which an electrical connector 24 links them to a row driver 26. Both the row driver 26 and the column driver 18 are connected to a controller 6 28 which will typically comprise a programmed microprocessor or an application specific integrated circuit (ASIC). Other electrode configurations can be applied to the liquid crystal device to provide, for example, a seven segment display, an r,0 display and so on. The liquid crystal device will contain a ferroelectric liquid crystal material such as SCES (Merck Ltd., Merck House, Poole U.K.) and will also comprise polarising means and alignment layers (not shown) as is known to those skilled in the art. Alternate electrodes on each wall of the device may be connected to the row and column drivers at opposite edges of the walls.

Each of the column electrodes of the liquid crystal array of figure 1 effectively comprises a large capacitance driven by a voltage source via a resistance. A lumped equivalent circuit is shown in figure 2 which shows an AC voltage Vsin cot applied across a resistor R and a capacitor C connected in series. It will be understood that the dissipation of power in the resistor is dependent upon the frequency 0) of the voltage applied. At higher frequencies the impedance of the capacitor is smaller causing a greater voltage to be dropped across the resistor R leading to a higher power dissipation. Similarly for higher frequencies a reduced voltage is present across the capacitor (i.e. the liquid crystal cell) reducing both contrast ratio and switching discrimination. The average power dissipated by such a circuit is given by /P) 0) 2c2v2 R 2(1+ (02 R 2C2) From this equation it can be seen that the power dissipation is heavily dependent upon the frequency co when driven by a sinusoid. When driven by a square wave the average power dissipation is:

= 4CV 2 /1.a.t.(1.a.t. represents the addressing speed and is explained below) 7 Figure 3 shows a distributed, or transmission line model, of a column electrode in which a plurality of series-connected resistance r are connected as a ladder in which a plurality of capacitors c comprise the rungs. The equivalent circuit is driven by a voltage Vsin ot as before. The average power dissipated in this arrangement is given by:

v 2 sin _2corc sinh J5rc (P)= 0 TC f 2 F2r 1 cosh -J-rc + cosVrc and the voltage drop along the electrode track is:

2 ex v P( v V1 + 2 exp(--..'2-o)rcos(-2rj)-rc)+ exp(2.v'2--o)rc) The power dissipation for this equivalent circuit remains the same as for the lumped circuit model when the electrode track is driven by a square wave, at:

= 4CV'/1.a.t.

From the equation for the voltage drop along the electrode track, however, it can be seen that the higher frequency components suffer a large reduction in voltage along the track. The power contained in these components is therefore dissipated at the beginning of the track which is at the edge of the panel. Consequently the edge of the panel gets hotter than the remainder of the panel. Even if the column electrode signals are applied at alternate edges of the panel a rather uneven temperature will exist over the panel. By reducing the amplitude of the higher frequency components in the column drive signals with respect to a square wave, the non-uniformity of heating is reduced.

8 Furthermore, it is known to reduce the effective resistance of the electrode tracks in a ferroelectric liquid crystal display by providing a low resistance element alongside the transparent electrode. Although such a low resistance element will not usually be transparent it can be very narrow and placed in the inter-pixel gap in a display. This can reduce the effective resistance of the track markedly. However, when the column electrodes are driven by square wave signals, there is no reduction in the heating of the panel because the dissipation of power for square wave drive is independent of the resistance (see equation above). However, as will be appreciated from the equation for power dissipation in the distributed circuit model, when a sinusoidal waveform is applied to the column electrodes, the resistance r does have an effect on dissipated power. Thus by applying the column waveforms in accordance with the invention to a liquid crystal array device having reduced resistance electrode tracks a reduction in the overall heating of the device can be acheived.

Figure 4 (a) shows a first drive waveform for the column electrodes of the present invention. The waveform is a sinusoid having a period of one line address time or 1.a.t. The Lax is the time that spent addressing a particular row of the display and in simple drive schemes is the duration of the strobe pulse applied to the row. However, more sophisticated drive schemes use a strobe pulse that overlaps for two adjacent rows (see, for example U.K. Patent number 2,262,83 1) so is it better to define the 1. a. t. as the frame time for addressing the whole array divided by the number of rows, thus:

1/(frame rate x number of lines addressed per frame) for a ferroelectric liquid crystal device this will typically be 25 pis or less.

The 1. a.t. shown in figure 4(a) is divided into two equal time slots and for this reason the driving scheme is known as a two slot scheme. More complex schemes, for example a four slot scheme, exist but for the sake of simplicity the present part of the discription will concentrate on a two slot scheme. The strobe signal will generally comprise an amplitude of zero in the first slot and a positive-going square wave in the second slot. The resultant waveform applied to those cells that are actually being addressed is the 9 combination of these two signals. Two data waveforms are required to provide a resultant signal that will cause the cell to change state and a resultant signal that will not cause the cell to change state. The two data waveforms are often the inverse of each other so figure 4(a) also shows a reversed sinusoidal waveform which comprises the other data (column) signal.

Figure 4(b) shows a pair of data waveforms having a triangular shape. In combination with a suitable strobe signal, one of these two waveforms wi11 cause the relevant cell to switch while the other waveform will leave the cell in its original state. Figure 4(c) shows a pair of data waveforms, again inverses of one another, based upon trapezoidal waveforms. All of these waveforms have a considerably reduced higher harmonic content compared with a square wave and so provide a more even heating of the display device panel.

Figure 5 shows a block schematic diagram of a driving arrangement 100 for applying data waveforms in accordance with the present invention. A ferroelectric liquid crystal array 102 comprises a plurality of columns numbered 1 to n of which numbers 1, 2, 3 and n are shown. The driving of the array is controlled by a clock generator 104 which governs the timing of the signals applied to the array. The clock generator 104 is connected to a row driver 106 which is connected to all of the rows of the array to provide the strobe signals at the correct time to the appropriate row.

The clock generator is also connected to a data source 108 which provides the data relating to the desired state of each pixel in a particular row for each application of the strobe signal. A signal from the clock generator 104 clocks this data into a shift register 110 every time that a new row is addressed. The shift register has n outputs Q 1 to Qn, in other words one for each column of the display, and each of these outputs controls one of n analogue switches 112. Under the control of the outputs of the shift register 110, the analogue switches couple either a SELECT or a NON-SELECT data signal to their respective columns of the array. The SELECT data signal is provided by a digital to analogue converter (DAC) 120 which is provided with digital data from a random access memory (RAM) 116. The NON-SELECT data signal is provided by a DAC 118 provided with digital data from a RAM 114. The RAM 116 and the RAM 118 contain digitised versions of the SELECT data and NON-SELECT data waveforms shown in figure 4. The RAMs are addressed by the clock generator 104 providing a parallel signal which counts up at a fast rate to clock the digital signals representing the data signals out of the RAMs. The DACs convert these signals into a pair of substantially continuously varying signals which are applied to respective poles of the switches 112. The relevant data waveform is selected from the outputs of the DACs by the plurality of switches 112 and the required combination of strobe waveform and data waveform can be applied to each cell in the array. The RAMs must be clocked at a sufficiently high rate and the RAM/DAC combination must be of high enough resolution to mimic the desired switching waveform accurately. Some examples of suitable circuitry are as follows. The RAM may comprise part number CY7C128-45PC from Cypress Semiconductor which provides 2k x 8bit of memory with an access time of 45ns. The DAC may comprise part number DAC08CP which has an 8 bit current output with an 85ns settling time although this may need a current to voltage converter. Alternatively, the DAC may comprise part number OPA 600 available from Burr-Brown which provides a +/- 1 Ov output and a settling time to 0. 1 % of 8Ons. This combination of circuitry will give 256 voltage steps and 100 time steps in a 10is time slot if it is clocked at 1ONffiz.

The row driver may be arranged to provide a bi-directional strobe or, alternatively, a blanking pulse ahead of the application of the strobe signal as is known in the art. The blanking pulse is chosen to switch the cells in a particular row into a given state regardless of the data waveform applied to the cell at that instant. As is known the blanking pulse allows the array to be driven using a strobe signal having a monopolar pulse. The blanking pulse is typically applied 5 to 10 rows ahead of the strobe signal.

Where the SELECT data waveform and the NON-SELECT data waveform are inverted versions of each other such as shown in figure 4 then the RAM 114 and the DAC 118 can be omitted. In this case the NON-SELECT waveform may be derived from the SELECT waveform by using an inverting buffer connected to the output of the DAC 120. Where the data source 108 can provide the required data in a parallel format, the shift register may be omitted and the data source connected to control the analogue switches 112 directly. The clock generator 104 may also be provided with means to alter the data waveforms in response to operational data from the liquid crystal device array. For example, it may be desired to change the amplitude and/or the shape of the data waveforms as the array becomes hotter in use. This can be readily achieved by providing the data corresponding to the further waveforms in the RAM and altering the addressing of the RAM to output the modified data waveform as appropriate.

It is also possible to provide the appropriate SELECT and NON-SELECT data waveforms by analogue means, particularly for the case of a sinusoidal waveform. One such circuit is a waveform generator integrated circuit part number ICL 8038 available from Harris Semiconductor. This can provide both sin and triangle waveforms from 0. 00 1 Hz to 1 0OkHz using voltage control of frequency. Using a digital signal generating arrangement as shown in figure 5, however, will generally be easier and more flexible.

Figure 6 shows a strobe signal and a pair of data signals in accordance with the invention. In this case the strobe signal has a positive-going voltage portion which occupies three slot widths, in other words it is one and a half times as long as the 1. a. t. The two alternative data waveforms are sinusoids occupying two time slots each and being inverted versions of each other. Although it is not strictly to scale, figure 6 also gives an impression of the relative amplitudes of the strobe and data waveforms.

Figure 7 shows a graph of the operating window of a ferroelectric liquid crystal array device driven using the waveforms shown in figure 6. The vertical axis indicates the switching time of the cells in the device measured as the slot width of the applied signals in microseconds. The horizontal axis is the applied peak strobe (row) voltage. The two curves on the graph are each associated with a diagram of a waveform which is the resultant signal applied to a cell. In the case of the curve identified using solid 12 squares it is the NON-SELECT waveform and the curve represents a suitable TV curve for NON-SWITC1UNG driving of the cell. The curve represented by hollow squares relates to the resultant waveform shown which is for SWIT0HNG driving of the cell. The curve represents a suitable TV combination for this waveform. This graph illustrates a good switch ing margin and discrimination between switched and unswitched states for ferroelectric liquid crystal device operation. For example if the panel is driven at 10pis slot width then operation with a strobe voltage between approximately 27 and 36 volt is possible, allowing for some variations (such as thickness or waveform distortion) over the panel area.

The present invention also provides an improvement in so-called pixelpattern dependent heating of the ferroelectric display device. This phenomenon is not widely recognised and so will be described briefly here.

When one of two data waveforms may be applied to address the successive rows of an array device, the signal applied to the column electrode will either be the same for addressing successive rows or it will change if the adjacent pixels in the column are in different states. So, if adjacent pixels in a column are all black (say) the waveform applied to the column will be a continuous sinusoid for the data waveforms shown in figure 4(a). Where the adjacent pixels are black, white, black, white and so on the data waveform applied to the column electrodes will invert for successive rows. Figure 8(a) shows a pair of data waveforms according to the prior art (square wave type) and figure ") and (c) show a pair of data waveforms applied to the column of a liquid crystal array when adjacent pixels in a column are black, black, black, black (say) and when adjacent pixels are black, white, black, white respectively. The latter waveform in figure 8(c) has double the wavelength of that shown in figure 8(b). From the discussion above regarding heating of a panel, it will be understood that the former waveform results in rather more power dissipation than the latter and hence more heating of the panel. Thus the panel heating depends on the pattern displayed, leading to pattern dependent heating.

13 Figure 9 however, shows the corresponding two waveforms for pixel patterns of B, B, B, B and B, W, B, W when the data waveforms comprise sinusoidal waveforms in accordance with the present invention. The first waveform is a pure sinusoid while the second waveform is sinusoidal in shape but inverts every 1.a.t.

The heating power of the two waveforms shown in figure 9 is almost identical and this can be confirmed as follows. The lower waveform shown in figure 9 is defined as:

g(x) = sin 2nx/L for 0 < x < L, g(x) = - sin 2nxIL for -L < x < 0 and the Fourier expansion for this waveform. is given by:

8 n= g(x) 2) cos L n=1,3, (4 - n 7C For comparison, we shall consider a square wave defined as:

g(x) = -1 for -L < x < -LI2, g(x) = 1 for -L/2 < x < 1/2, g(x) = -1 for L/2 < x < L whose Fourier expansion is given by:

4 00 1 n= gX) = - cos 7r 1 n=1,3,5 n L So it will be appreciated that the amplitude coeffilcients of the lower waveform shown in figure 9 decrease far more rapidly than those of a square wave. In other words, in the lower waveform shown in figure 9, the power is concentrated into the lowest frequency components.

14 Thus, for sinusoidal data waveforms (and also for the waveforms of figures 4(b) and (c) but slightly less so) the problem of pattern dependent heating is considerably reduced.

Figure 10 shows a block schematic digrarn of a test rig to test the teachings of the present invention using a ferroelectric liquid crystal array device. A sinusoidal data waveform at a frequency of 10 kHz was applied to all of the column electrodes while the row (strobe) electrodes were grounded. A number of temperature measurement points were established substantally along a centre line between the strobe electrode attachments at progressively greater distances from the edge at which the data signal was applied. A square wave of the same frequency and the same rms voltage was applied to the test rig for comparison purposes. The results, measured after the panel had come to equilibrium, are shown in figure 11.

Figure 11 shows a graph of temperature increase in degrees centigrade on the vertical axis against distance from the driven edge of the panel for the two waveforms. The sinsoidal waveform gave the temperature effects shown in the curve having a number of filled circles and the square waveform gave the temperature effects shown in the curve having a number of filled squares. For the edge of the panel, the square waveform resulted in a temperature rise of nearly double that resulting from the sinusoidal waveform. For the temperature-sensitive ferroelectric liquid crystal display panel this is particularly significant.

As mentioned previously, waveforms suitable for use in the present invention may be defined in terms of their power dissipation or their waveform distortion. Considering power dissipation, where low distortion is assumed, the power of a sinusoidal waveform is of the form: = C02C2 v 2 R while for a square wave the power is of the form:

= COCV2 where C is the panel capacitance, R is the sheet resistance of the column (data) electrodes and V is the amplitude of the data waveforms. These equations can be combined to give a generalised approximation to the power of a waveform as:

= CV 2 o) (RCco)n where n = 0, 1 are the sin and square wave limitOther waveforms such as triangular and so on will have values of n somewhere between these two limits. Another parameter that affects the heating performance of the data waveforms is the number of slots m in the data pulse. The panel under consideration has a diagonal of 1 metre. Using these parameters, a suitable data waveform would satisfy the inequality:

[CV2m/(2 1.a.t.)][7cRCm/(2 1.a.t.)]n < 100forO<n< 1 while a waveform satisfying the inequality:

[CV 2 m/(2 1.a.t.)][nRCm/(2 1.a.t.)]' < 50 forO<n< 1 gives improved performance.

When the data waveforms are sinusoidal, heating performance is satisfactory when the inequality:

C2VIRM. 2 /1. a. t.' < 40 is satisfied, while improved performance will result if the inqualty:

C 2 V 2 Rm2/1.a.t.2 < 20 16 is satisfied.

When the distortion of the waveforms along the column electrodes of the array are considered, if the data waveforms comprise sinusoidal waveforms, the following inquality should be satis fied:

CRm/(2 1.a.t.) < 0.25 where the parameters are as defined above. Improved performance will result if the following inequality is satisfied:

CRm/(2 1. a. t.) < 0. 15 which is particularly significant if the lowest possible 1.a.t. is to be used with a ferroelectric liquid crystal display panel (for the fastest possible addressing). When the 1.a.t. is longer than the minimum possible value then the effect of waveform distortion on performance of the panel becomes less significant. The above restrictions on waveform distrotion will be particularly significant is the Lax is reduced below I OPLs (for example 7.5gs) for a large area ferroelectric liquid crystal device panel.

The present invention is not limited to those aspects described above but also encompasses improvements, variations and further refinements as will be apparent to a person skilled in the art.

17

Claims

CLAIMS:

1. A ferroelectric liquid crystal device comprising a layer of ferroelectric liquid crystal material contained between a pair of walls and a first plurality of electrodes and a second plurality of electrodes defining a plurality of addressable liquid crystal cells and a driving arrangement for applying a first signal in succession to thefirst plurality of electrodes and for applying a plurality of second signals simultaneously to the second plurality of electrodes, wherein the plurality of second signals comprise non-square wave signals which have a lower harmonic content than a square wave.

2. A ferroelectric liquid crystal device as claimed in claim 1, wherein the plurality of second signals have no effective harmonic content above the fifth harmonic of the fundamental.

3. A ferroelectric liquid crystal device as claimed in claim 1 or claim 2, wherein the second signals comprise signals having a substantially continuously varying level.

4. A ferroelectric liquid crystal device as claimed in claim 1, claim 2 or claim 3, wherein the second signals comprise sinusoidal signals.

5. A ferroelectric liquid crystal device as claimed in claim 1 or claim 2, wherein the second signals comprise triangular signals.

6. A ferroelectric liquid crystal device as claimed in claim 1 or claim 2, wherein the second signals comprise trapezoidal signals.

7. A ferroelectric liquid crystal device as claimed in claim 1, wherein the inequality: [CV'm/(2 1.a.t.)][nRCm/(2 1.a.t.)]n < 100 forO<n< 1 is satisfied and in which C is the device capacitance, V is the amplitude of the second signals, m is the number of slots in the second signals, 1.a.t. is the line address time of 18 the device, R is the sheet resistance of the second plurality of electrodes and n is a parameter relating to the shape of the second signals as defined herein.

8. A ferroelectric liquid crystal device as claimed in claim 7, wherein the inequality: [CV2 m/(2 1.a.t.)][7cRCm/(2 Lax)]n < 50 forO<n< 1 is satisfied.

9. A ferroelectric liquid crystal device as claimed in claim 1, wherein the second signals comprise sinusoidal signals and the inequality: c 1V2RM2/1 a. t.2 < 40 is satisfied and in which C is the device capacitance, V is the amplitude of the second signals, R is the sheet resistance of the second plurality of electrodes, m is the number of slots in the second signals and Lax is the line address time of the device as defined herein.

10. A ferroelectric liquid crystal device as claimed in claim 9, wherein the inequality: c 2VIRM2/1 a. t.2 < 20 is satisfied.

11. A ferroelectric liquid crystal device as claimed in claim 1, wherein the second signals comprise sinusoidal signals and the inequality: CRm/(2 Lax) < 0.25 is satisfied in which C is the device capacitance, R is the sheet resistance of the second plurality of electrodes, m is the number of slots in the second signals, 1.a.t. is the line address time of the device as defined herein.

12. A ferroelectric liquid crystal device as claimed in claim 11, wherein the inequality: CRm/(2 Lax) < 0. 15 19 is satisfied.

13. A ferroelectric liquid crystal device as claimed in any one of the preceeding claims, wherein the device comprises a large area ferroelectric liquid crystal display device.

14. A driving circuit for a ferroelectric liquid crystal device which device comprises a matrix of liquid crystal cells addressable via a plurality of row electrodes and a plurality of column electrodes, the driving circuit comprising row driving means for applying a first signal in succession to the plurality of row electrodes and column driving means for simultaneously applying a plurality of second signals, which second signals each comprise one of at least two data signals, to the plurality of column electrodes, wherein at least the means for applying a plurality of second signals provides a signal, at least a portion of which signal has a substantially continuously varying level.

15. A method of driving a ferroelectric liquid crystal device which device comprises a matrix of liquid crystal cells addressable via a plurality of row electrodes and a plurality of column electrodes, the method comprising driving the rows of the device by applying a first signal in succession to the plurality of row electrodes and driving the columns of the device by simultaneously applying a plurality of second signals to the plurality of column electrodes, which second signals each comprise one of at least two data signals, wherein at least a portion of the data signals has a substantially continuously varying level.